READING COMMERCIAL LIABILITY INSURANCE POLICIES (like you ve never read them before) 1

READING COMMERCIAL LIABILITY INSURANCE POLICIES (like you ve never read them before) 1 By R. Lee Akazaki, C.S. Partner, Gilbertson Davis Emerson LLP Past-President, Ontario Bar Association (a) (b) (c) Your client has been sued for tort damages. Your client has sued for tort damages. You are acting for an insurer whose policyholder has been sued. But wait. Is the claim one for which liability coverage will respond? Will the insurer defend the lawsuit? Will there be indemnity after judgment? Will there be a settlement? Despite the frequency with which one or more of these basic litigation scenarios arise in practice, no two cases are exactly alike. As practitioners, we are also prone to leaping to conclusions without engaging a step-by-step coverage analysis. It does not help that the judicial interpretation of liability insurance coverage exclusions consists of generalities which the courts often fail to apply with systematic consistency. Nevertheless, a bit of patience goes a long way to make sense of the insurance coverage on any given claim. Facts for Interpretation of Liability Insurance Policies The coverage exercise should always start with the applicable facts. It is an error to read the insurance policy first. Reading the terms of coverage or exclusions from coverage 1 Originally Presented to the Canadian Bar Association Yukon Branch, July, 2011; Revised December, 2011

- 2 - will import bias in favour of your client, whether insured, insurer, or party plaintiff. The judge dealing with the coverage issue will start with the facts, then the insurance contract. So should you. It is beyond the scope of this paper to detail the law relating to the duty to defend and the duty to indemnify, or the ins and outs of insurer strategies such as non-waivers and reservations of rights, but I would be happy to answer any questions. However, the essential functions of liability insurance served by defence and indemnity will determine the nature and source of the facts which start your inquiry. This dichotomy exists in the law of insurance, because liability insurance policies are structured to provide these coverages separately. The duty to defend focuses on the facts of a claim for liability at the beginning of a lawsuit, at the pleadings stage. The duty to indemnify will be determined by facts at the end, as a determination by the court adjunct to the judgment; or in some cases after settlement, after which the insured might bring a suit for wrongful denial of coverage. The scope of the duty to defend is defined by the terms of the coverage afforded under the policy. What does this mean? Ultimately, the inquiry requires a choice of facts within a continuum between the hypothetical (the pleadings rule) and the definite (the judgment rule). Duty to Defend The facts are those pleaded in the statement of claim or other statement defining the litigated dispute. The widest latitude is afforded to the reading of the pleadings and to the terms of possibility of coverage under the policy. Nichols v. American Home Assurance

- 3 - Co., [1990] 1 S.C.R. 801, is the seminal decision, both for the duty to defend and the duty to indemnify. Duty to Indemnify Judgments The facts are those determined by the court or other tribunal after evidence is presented at the trial or other substantive hearing. Unlike the duty to defend, the duty to indemnify places the insurer defending a suit in the position of oracle, divining the future resolution of an action at trial. This task requires constant reassessment of the evidence as a matter progresses through the discovery stage of a lawsuit. The factual outcome will only be fixed at trial. Where an insurer has agreed to defend the suit but has reserved its rights on coverage, it is possible in some instances for a conflict of interest to occur. Such instances may affect the right of the insurer to appoint and instruct defence counsel. Duty to Indemnify Settlements What facts apply where a coverage dispute involving liability insurance survives the settlement of a lawsuit? Unlike the determination of fact available to a trial judge after a judgment or verdict has been made in the underlying suit, settlements are made on contingent facts in that the insured defendant will most likely continue to deny the facts giving rise to liability. Nevertheless, settlements are made based on the evidence available at that fixed moment in time, not on bald allegations. Insurers occasionally argue there can be no coverage where the facts alleged by the plaintiff are strongly denied by the defendant and, on the available evidence, patently untrue. Such an approach is usually incorrect.

- 4 - Most policies of liability insurance contain language requiring the insurer to defend even if the action is groundless. Cases proceeding to judgment are also assessed for coverage based on a contingent set of facts and denial by the defendant. If this is so for cases proceeding to judgment, why would the principle be any different in the case of settlements? A settlement proceeds along the same litigation path, and the only difference is that the path is arrested at some time before judgment is handed down. Moreover, a party abandoned early on by a liability insurer can also settle a suit whose principal exposure is legal costs, provided a court would consider such a settlement reasonable in the circumstances. The key to solving this puzzle is the position of the parties after the defence obligation is assessed by the insurer. Ordinarily, a coverage dispute arising from a settlement will come about after an insurer has declined coverage and is not defending. The insured will settle the claim and seek indemnity from the insurer. (Where an insurer has agreed to defend, either outright or under some reservation of rights, it is almost invariably the insurer who will settle the action. This is a function of provisions in most liability policies prohibiting insureds from accepting fault or settling claims without the consent of the insurer, the consequence of such conduct being the voiding of liability insurance coverage.) The settlement conundrum was squarely addressed in the case of Cansulex Ltd. v. Reed Stenhouse Ltd., 1986 CarswellBC 290, 70 B.C.L.R. 273, 18 C.C.L.I. 24, [1986] I.L.R. 1-2090, at paras. 195-96: The liability of Cansulex has never been determined judicially and the insurers should not be condemned to pay any amount except in

- 5 - accordance with their policy obligations unless the law provides otherwise. On the other hand, the law would fit Mr. Bumble's characterization if, after a denial of coverage, Cansulex is required to prove itself liable to Bibby in order to recover a reasonable amount paid to buy peace when it has spent 33 days and over $2,000,000 in a foreign court asserting the opposite. In my view, an insured in the position of Cansulex is entitled to settle any insured claim brought against it on any reasonable basis and for any reasonable amount, and to recover such amount from its insurers not necessarily as indemnity, as such may only be payable in discharge of a liability which may not have existed, but as damages for breach of contract. In this connection it is my view that an insured who has been abandoned by his insurer is entitled to buy peace at a reasonable price even though he denies his liability. The fact that he pays the claimant $"X" and pays his solicitors $"Y" does not make the two payments intrinsically different. They are amounts he must pay in order to defend and protect himself from greater loss or expense. These amounts, if reasonable, are payments which the insurers might themselves have paid if they had not wrongly repudiated their liability to the insured, and they cannot insist as a condition of indemnity or reimbursement that the insured stubbornly maintain a denial of liability all the way to judgment in order to obtain the protection of the policies which has been wrongly denied. The principle in Cansulex has been met with approval in Ontario: John Picken Ltd. v. Guardian Insurance Co. of Canada (1993), 17 C.C.L.I. (2d) 167 (Ont. C.A.), at paras. 16-17: 16. The right of the respondents to recover the settlement funds that they paid to settle the Knoch estate action can be justified on two bases. First, a reasonable settlement can be viewed as a foreseeable consequence of the appellant's breach of contract by its wrongful refusal to defend the Knoch estate action. This was the effect of the judgment in Cansulex Ltd. v. Reed Stenhouse Ltd., [1986] I.L.R. 1-2090 (B.C. S.C.). We agree with the approach taken by McEachern C.J. in that case. We do not hold that any settlement is reasonably foreseeable. For the purposes of this case we need go no further than to say that this settlement (the reasonableness of which is not in issue), made on the advice of counsel, is a reasonably foreseeable consequence of the appellant's breach of its duty to defend the respondents, Jon Picken Limited and Jon Picken.

- 6-17. Second, we note that the policy issued by the appellant defines "loss" as referred to in the policy, to include a settlement. It distinguishes between a settlement and a judgment, providing indemnity for both subject, of course, to the coverage provisions of the policy. It would follow, therefore, that under the terms of the policy itself, the appellant is contractually obligated to indemnify the respondents Jon Picken Limited and Jon Picken for the amount that they paid to settle the Knoch estate action. An appreciation of the case law in Canada and in the United States tends to reveal two general statements: 1. Where the insurer did not defend and is found to have denied coverage wrongly, the insurer will be held liable for a reasonable settlement entered into by the insured. The court will not require the insured to prove liability to the plaintiff, in situations where the insured defended on the basis that the plaintiff s claim had no merit. 2. Where the insurer defended and reserved rights to deny coverage, the court will look to the evidence available to the parties at the time of settlement. The standard appears to be the facts reasonably informing the settlement, i.e. the decision to pay. Decoding Coverage under a Liability Policy Insurance as Contract Interpretation with and without Ambiguity The conventional starting point for interpretation of coverage under an insurance policy was restated by the Supreme Court of Canada in Canadian National Railway Co. v.

- 7 - Royal and Sun Alliance Insurance Co. of Canada, 2008 SCC 66, [2008] 3 S.C.R. 453 (CNR). The first step is to apply the general law of contract. The court must give effect to the intention of the parties to be gathered from the words they have used. If there is no ambiguity, then no special rules of construction are required. The rules of construction are intended to resolve doubt in favour of one party over another when ambiguity exists, not to create ambiguity where there is none and where the ordinary rules of construction can be used. As stated by Justice Binnie, at para. 33 of the CNR decision, The expression faulty or improper design requires interpretation but I do not think it is ambiguous, meaning that the judicial act of interpretation involves applying the words in the contract to the facts. When seen in this light, ambiguity is not the same as requiring thinking or hard work. Ambiguity means capable of having two or more relevant meanings, one being applicable and the other not applicable. In this regard, both the majority and the dissent in CNR were in agreement. 2 This important point is perhaps the most frequently overlooked in practice, and the courts have not always seized on it. Lawyers and judges are frequently rushing to the special rules of construction when no need arises, the policy wording being clear enough for most reasonable people to comprehend its intention. Moreover, even unclear or badly worded clauses are not necessarily ambiguous. An alternative meaning must also be capable of giving effect to the parties reasonable expectations. In the absence of at least one other such meaning, there can be no ambiguity. Perhaps the best way to view the rules of construction are like the presumptions used in organized sport to resolve the limits of human ability to adjudicate eg., the tie 2 see CNR, at paras. 30-33 (per Binnie J.) and 73-76 (per Rothstein J.)

- 8 - goes to the runner. If there is no tie, then the runner is safe or out, and there is only one reasonable interpretation. No special rule is required. To continue the sporting analogy, a close play does not justify the invocation of a rule. A close call does not require resolution of an ambiguity. Rather, only where the umpire or referee cannot tell which play occurred first is any rule required. True ambiguity exists only when the ball reaches first base at the same time as the runner, or in tennis when the ball hits a line. In those instances, a rule replaces the factual interpretive process. It is the same for insurance policy exclusions. Where we often encounter difficulty is the juxtaposition of clauses defining insured risks, compared to those defining exclusions. Many judicial authorities and texts refer to the Supreme Court s decision in Reid Crowther & Partners Ltd. v. Simcoe & Erie General Insurance Co., [1993] 1 S.C.R. 252, 99 D.L.R. (4th) 741 at p. 752. There, general principles were stated as including: (1) the contra proferentem rule, that any ambiguity in the document will be construed against the party who drafted it; (2) the principle that coverage provisions should be construed broadly and exclusion clauses narrowly; and (3) the desirability, at least where the policy is ambiguous, of giving effect to the reasonable expectations of the parties. This badly-ordered formulation has led to the widely-held presumption that all insurance contract provisions are ambiguous, an untended consequence of unclear editing.

- 9 - In fact, ambiguity is a precondition for the application of the test. The test does not apply in the absence of ambiguity. In terms of priority, the restatement in Reid Crowther was also expressed in the wrong order. It should have been (3),(2) then (1) from preserving the intention of the parties, to the implied reading of coverage over exclusions, and only then to the power imbalance between insurer and insured. The Supreme Court in CNR has helped to right the order of operations problem, but judges may still be tempted to use judicial onuses and presumptions as crutches on which to lean. We as counsel are charged with the task of educating the court to navigate through this exercise. A further subtlety in the interpretive technique arises exclusively in liability insurance. Increasingly, liability insurance decisions arise in the context of duty to defend, as opposed to indemnity. This trend may be influenced by mounting litigation costs, where the duty to defend a claim with questionable merit or an inflated claim for damages may be more valuable to the insured than the contract of indemnity. Between cases involving defence and indemnity, it can be easy to conflate contract law s presumption against ambiguity of meaning in the policy wording (i.e. we must strive to give effect to the parties intentions) and the presumption of possible outcomes when reading a plaintiff s pleading (often expressed in terms of ambiguous pleadings). In this regard, it is important to note that in duty to defend cases, it is not necessarily policy interpretation but rather pleading interpretation which causes insurers problems. A notable example of the often misunderstood subtlety appears in popular reaction to the recent Supreme Court decision in Progressive Homes Ltd. v. Lombard General Insurance Company of Canada, 2010 SCC 33. Progressive Homes is considered the current

- 10 - law on insurance coverage interpretation. Careful reading of the decision reveals that, throughout, the court relied entirely on a plain-language analysis. It tried to stay away from having to resolve any ambiguities, and only mentioned the interpretive principles for resolution of ambiguities once in obiter as a hypothetical (para. 57). Throughout the reasons, at paragraphs 15, 35-37, 44, 49 and 56, the court relies on the plain meaning of the policy wording. In other words, there is no judge-made law in operation here. Read properly, Progressive Homes cannot be considered inherently pro-insured or pro-insurer. Instead, the court s references to ambiguity favouring the insured all relate to the pleadings (paras. 51, 54, 70 and 71). The court, in reviewing the pleadings generously and capable of various trial outcomes, applied the threshold for duty to defend. The statement, Lombard must show that an exclusion clearly and unambiguously excludes coverage, arises from the notion that the scope of the duty to defend is defined by the possibility of coverage on the pleadings (para. 51). Once one appreciates that the court is resolving ambiguities on the pleadings and not on the insurance contract, it is possible to see that Progressive Homes did not chart a new course at all. Rather, the facts were different than in Swagger Construction Ltd. v. ING Insurance Co. of Canada (2005), 47 BCLR (4th) 75, which some have incorrectly said Progressive Homes reversed. Swagger dealt with a delayed counterclaim for repair of construction deficiencies in response to the contractor s own claims for economic losses. The nature of the remedial work resulting from construction defects were generally limited to the defective structural elements including the building envelope (paras. 14-16). A very small portion of the counterclaim dealt with property damage to features other than the

- 11 - building envelope, such as carpeting and drywall, which the court did consider as possibly coming under the definition of property damage (paras. 43-46). Although the Supreme Court in Progressive Homes implied that Swagger limited property damage under a CGL policy to damage to third-party property, Swagger in fact did not really deal with damage caused by the generally defective building envelope, but rather the repair of the building envelope itself. In contrast, Progressive Homes involved damage caused by water leakage due to faulty installation of various weatherproofing features of buildings. At paragraph 38 of Progressive Homes, the court noted that the repair of a defect is not included in the definition of property damage, and this was conceded by the insured, Progressive. The contentious aspect of Progressive Homes was not the repair of defects but rather the repair of the effects of water damage. In particular, there were significant pleadings of deterioration of building components, moisture penetration, rot and infestation. Although the insurer Lombard argued that the damage was not an accident and therefore did not meet the fortuity element of insurance, the Supreme Court of Canada did not agree that the damage was not accidental (para. 50). Nor did the court hold much stock in the notion that the CGL policy would be transformed into a performance bond, because the work had already been completed (para. 48). Where the court could have been clearer in its reasoning was the notion that defective construction could be accidental in nature. At paragraphs 45 and 46, the court did acknowledge that defective workmanship has been held to be non-accidental. Generally speaking, the failure to construct a building with required materials or building practices is a deficiency issue and is not insurable. A

- 12 - distinction must be made, however, between the defect (non-accidental) and the consequence of the defect (accidental). Reading the Progressive Homes decision as a whole, and in view of the facts, it was only logical that Lombard s argument would fail. As stated later in this paper, courts determination of duty to defend resolve ambiguities in pleadings (as opposed to policies) against insurers. Consequently, at paragraph 46, the court ruled in favour of the insured on this point because it could not agree with Lombard s view that faulty workmanship is never an accident. The italics for the word never are in the judgment itself, an indication of the low threshold on a duty to defend application. Had the claims involved recovery of the cost of repairing leaks or other defects to prevent anticipated water infiltration and consequential loss, the Supreme Court would likely have decided Progressive Homes differently. The element of fortuity is lacking in pure deficiency-repair litigation, because repair or replacement of defective construction is inevitable because the defect must be corrected (para. 38). The causal connection involving risk simply does not exist. In contrast, water entering a building structure is not a defect but a consequence of a defect. The damages in that instance go beyond correcting the defect but extend to consequential damage caused by a factor external to the defect. What appears to have been missing from Progressive Homes (and from Swagger, in some respects) was any discussion of the principle of apportionment. In a context where the insured had conceded lack of coverage for liability for making right the defects themselves, as opposed to the consequential water damage, imposition of the entire defence on the insurer would have been contrary to the apportionment principle. That

- 13 - apportionment principle is that the insurer should only pay for the defence it has contracted to provide: Hanis v. Teevan (2008) 92 OR (3d) 594 (Ont. C.A.), at para. 41; also Dunn v. Chubb Insurance Company of Canada, 2009 CanLII 7083 (ON SC). Returning to the test for interpreting insurance contracts, Progressive Homes actually reinforces a more balanced approach correcting the Reid Crowther mistake. At paragraphs 22-24, the court restated more or less correctly the correct order, ranging from straight application of unambiguous language, to construing ambiguities in line with reasonable expectations, to the use of the contra proferentem onus. Although it did not factor in the result, at paragraph 57, the court referred to the contra proferentem principle. Had there been any ambiguity in the language of clause (Z), interpretive principles lead to the same result. This verges on circular reasoning, because any clause already read to be unambiguous would, almost inevitably, lead to the same result after the interpretive principles are invoked. Here, the court had, in paragraph 56, already relied on the plain language and found the exclusion language to be unambiguous. Anyone reading the phrase, by or on behalf of the Named Insured in the former clause can read that it is different from the new clause, which read: by the Name insured. By invoking contra proferentum, the court may have given the impression that it applied in this instance. Logically, it did not apply. Progressive Homes has also been construed in some quarters as a broadening of property coverage under commercial liability insurance, probably because of the belief that it reversed the reasoning in Swagger. (Eg., Bulldog Bag Ltd. v. Axa, 2011 BCCA 178, at paras. 1 and 19) In particular, the focus is on Lombard s failed argument that property

- 14 - damage is limited to third-party damage and does not include damage to the insured s own work. In other words, Lombard argued the coverage cannot mean damage caused by other parts of the same building, where the entire property was constructed by the insured. Some practitioners have argued that the decision swept away prior court decisions restricting the scope of damages to third-party property. This is incorrect. The reason Lombard s argument failed was set out in paragraph 37 of the decision: Qualifying the meaning of property damage to mean third-party property would leave little or no work for the work performed exclusion. The Progressive Homes case involved British Columbia leaky condo litigation, where home purchasers sued the developer for water damage. Given that there has never been a limitation of the definition of property damage to third party property, Lombard s argument was simply a conflating of the coverage and the exclusion. Progressive Homes has been construed in many ways, but it should not be. It does, actually, provide much-needed guidance to insurance lawyers and their clients who tend to focus on exclusions as a starting point (as did Lombard s, apparently). As stated at the outset of this paper, reading the policy first and considering the facts second, leads to a haphazard interpretive approach. The lawyer or judge starts with the wrong threshold question, by asking what triggers the exclusion. This focus on the exclusion trigger as a point of departure is usually wrong. In liability policies, exclusions generally do not exclude liability triggers. They exclude liability for damages caused by those triggers. To read an exclusion correctly, one first has to ask whether the claim comes within the grant of coverage, and then one asks what part of the

- 15 - grant of coverage to which it applies. Simply put: What damages does the exclusion exclude from the liability insurance? Justice Rothstein, in Progressive Homes, stated: [26] The insurance contracts in this appeal are CGL policies. CGL insurance policies typically consist of several sections.... The policy will set out the types of coverage contained in the agreement, for example, property damage caused by an accident. [27] This is typically followed by specific exclusions to coverage. Exclusions do not create coverage they preclude coverage when the claim otherwise falls within the initial grant of coverage. Exclusions, should, however, be read in light of the initial grant of coverage.... [28] A CGL policy may also contain exceptions to exclusions. Exceptions also do not create coverage they bring an otherwise excluded claim back within coverage, where the claim fell within the initial grant of coverage in the first place. Because of this alternating structure of the CGL policy, it is generally advisable to interpret the policy in the order described above: coverage, exclusions and then exceptions. (emphasis added; references removed for ease of reading) Progressive Homes v. Lombard, paras. 26-28 A exclusion therefore operates as a derogation from the initial grant of coverage. It is not like a condition, the breach whereof the policy may be rendered null and void. If one follows Justice Rothstein s order of interpretation, one then starts with the right question: i.e., what kind of liability for damages, otherwise covered by the grant of coverage, is being excluded from the liability insurance? Although it is rarely (if ever) expressed, this basic structure of insurance grants, exclusions from coverage, and exceptions from exclusions, is

- 16 - the underlying logic behind the interpretive principle that exclusions are to be read more narrowly than coverage provisions. (More on this topic later.) One then turns to the work performed exclusion in Progressive Homes, and finds that Lombard faced two obstacles to having the court agree with its position: the plain language of a wording change limiting the scope of the exclusion to work performed by the Named Insured, and an subcontractor exception to the exclusion in the second version of the policy. Although the court was not required to rule on the point, it was clearly disposed to uphold Lombard s position, had the exclusion followed the previous wording ( by or on behalf of ) and/or had it not contained a subcontractor exception. The court read the change of wording in the first version of the policy as being similar to a subcontractor exception. Unless a property developer can say it built the entire home by itself, without subcontractors, it is hard to see how the exclusion in Lombard s policy could have been construed any differently, especially on a duty-to-defend application, in which the insured s position must be more or less certain. Review of the court s rationale could therefore lead to at least one outcome different from the one in Progressive Homes. Wording such as property damage to work performed by or on behalf of the Named Insured, without a subcontractor exception, could exclude liability for water damage resulting from allegedly defective construction, even though the damage to the defective part of the building (eg. the hole, crack or other building compromise) leads to damage to non-defective parts (eg., rest of the building structure, as well as non-structural elements such as flooring). Although such consequential damage would have been covered in the grant of coverage for property damage, the coverage would be taken away by an exclusion which adequately

- 17 - describes the excluded property as the work of the insured and the work on behalf of the insured. Resolution of Ambiguities Genuine ambiguities do arise from time to time, because of the reliance on standardized forms, but also often in manuscript insurance forms intended to customize insurance for the needs of a particular insured. What, then, does the law require the court to do in the instance of ambiguity? The insured has the onus of proving the loss came within the terms of an insured risk. The insurer bears the onus of proving it came within the terms of an exclusion clause. Many jurists fail to recognize that this dichotomy is simply an application of the general burden of proof in any civil proceeding, that the party advancing a proposition relating to a contract bears the burden. It is not a special rule applicable to insurance policies only. The first requirement is to give effect to the reasonable expectations of the parties. The courts most often cite this passage by Estey J. in Consolidated-Bathurst Export Ltd. v. Mutual Boiler and Machinery Insurance Co., 2008 SCC 66 (CanLII)[1980] 1 S.C.R. 888, at p. 901: Even apart from the doctrine of contra proferentem as it may be applied in the construction of contracts, the normal rules of construction lead a court to search for an interpretation which, from the whole of the contract, would appear to promote or advance the true intent of the parties at the time of entry into the contract. Consequently, literal meaning should not be applied where to do so would bring about an unrealistic result or a result which would not be contemplated in the commercial atmosphere

- 18 - in which the insurance was contracted. [underline and bold added to the word apart ] Generally speaking, the true intent of the parties means no more than to provide all-risk insurance on a business. In that context, an interpretation which means no coverage can ever be provided would have been against the business interests of the insured, and would have made no sense. Otherwise, the parties are free to expand or to limit the scope of coverage by negotiation. Where there is little or no negotiation, there is an expectation that the insurance will cover any aspect of the business which is not excluded. Perhaps the most subtle and least understood distinction is between the next two principles cited in Reid Crowther: (1) the broad-coverage narrow-exclusions principle and (2) the contra proferentum rule. The courts in Progressive Homes, at para. 24, as repeated in Bulldog Bag Ltd. v. AXA, 2011 BCCA 178, at para. 20, made the error of making the former simply a corollary of the latter. They are not two sides of the same coin. As illustrated in paras. 32-33 of the CNR decision, the Supreme Court correctly held that coverage provisions are to be read broadly and exclusions narrowly, and then stated that there are cases where contra proferentum cannot apply. The latter does not apply to manuscript policies or standard policies if terms are negotiated between sophisticated parties. The operation of one without the other proves, logically, that there are two independent principles at work. As stated earlier in this paper, the broad-coverage, narrow-exclusion principle is a function of the structural organization of the insurance policy as a contract. Exclusions are narrower than coverage provisions, even if they appear to cover similar damages or causes

- 19 - of damage, because exclusions are secondary logically to coverage. 3 Coverage is determined by the grant, and exclusions carve out from that grant. Exclusions do not exist outside of coverages. (The true corollary to that statement is the notion that the absence of exclusion does not create coverage: Progressive Homes, para. 27.) The most tangible illustration of this broad-narrow approach at work appears at paragraph 37 of the Progressive Homes decision: Qualifying the meaning of property damage to mean thirdparty property would leave little or no work for the work performed exclusion. 4 This interpretive principle is not, as many have approached it, based on a need for a rule which is onerous on insurers. Rather, it is a function of setting a boundary within a boundary. Contra proferentum is a contractual principle taken from the law of contracts of adhesion. Usually in consumer contracts, ambiguities are resolved against parties who drew the wording. In traditional contract law, it is not really an interpretive principle but rather a method of relieving a harsh consequence of contractual exemption clauses drafted in fine print. The most significant practical difference between the broad-narrow construction principle and contra proferentum is that, in the latter case, the court can simply render a contractual term unenforceable. In the case of most liability policies, such as home, automobile, and standardized commercial general liability, contra proferentum is used frequently to resolve ambiguities in favour of insureds. A good example of this judicial technique occurred in the recent Court 3 Consequently, an exclusion for heavy vehicles would likely be construed as restricted to large commercial vehicles and would not apply to passenger vehicles, in comparison with a grant of insurance for automobiles, even though all automobiles by their nature tend to be heavy. 4 Indeed, the true issue in Progressive Homes, as far as the case was concerned, was the scope of the exclusion. Had the exclusion been worded to include work performed by others, there may have been a different result.

- 20 - of Appeal for Ontario decision in AXA Insurance (Canada) v. Ani-Wall Concrete Forming Inc. (2008), 91 O.R. (3d) 481. At paragraphs 28-29, the court applied the judicial rules despite their acknowledgment that it may lead to a result contrary to that contemplated by the insurer in drafting the policy wording: [28] Ani-Wall submits that when the definitions of Property Damage and Ripping Tearing expenses are inserted into the Rip and Tear exclusion, the clause becomes incomprehensible and yields the following result: This insurance does not apply to any liability for physical injury to tangible property, including all resulting loss of use of that property, or loss of use of tangible property that is not physically injured for the actual expenses incident to the intentional destruction and removal of concrete products which are found to be defective. [29] AXA acknowledges that the Rip and Tear clause is badly drafted and that read literally, it is difficult to comprehend. AXA nonetheless urges a less-literal interpretation and submits that when the clause is read purposefully, its meaning is plain and obvious AXA will not indemnify Ani-Wall for the cost of tearing down and removing the defective footings and foundation walls. [30] AXA s proposed interpretation is not illogical. It presumably reflects the limitation on coverage that AXA sought to achieve. But AXA cannot get out from under the wording it chose to use, at least not without having this court rewrite the clause. That is not our function. The lesson from AXA is fairly simple. If an insurance policy exclusion is difficult to read, it may not be enforced. From a public policy perspective, this judicial rule is highly problematic, since the written document is proof that an exclusion was within the contemplation of the parties. While it is clever for the court to say its role is not to rewrite

- 21 - policy provisions, that is precisely what it is doing when it is effectively writing the exclusion out of the policy wording for ambiguity. It is the freedom to contract turned on end. This is, effectively, the unintended consequence of rushing to the contra proferentum rule before considering first the reasonable expectations of the parties, then the structural nature of grants of coverage being broader than exceptions. The rip-and-tear exclusion in the AXA decision was not, at the end of the day, that hard to read, in the context of insuring a provider of concrete building materials. In time, resort to principle will likely result in the law of insurance returning to the same law which applies to all written contracts. Until then, lawyers acting for insureds are obliged to play the ambiguities much in the way criminal defence lawyers practice in the zone occupied by the burden of proof. For their part, defence lawyers should not be so ready to give up on a reasonably understood insurance exclusion. Concurrent Causes Giving Rise to a Claim The rules for interpreting exclusions will certainly test the lawyer s skill when the facts suggest two or more concurrent causes giving rise to a plaintiff s damages. At one time, the conventional wisdom was that the court should consider the dominant cause of an accident. If that cause triggered an exclusion, there would be no coverage. This proposition was held to have no application in the law of Canada, in Derksen v. 539938 Ontario Ltd., 2001 SCC 72, [2001] 3 S.C.R. 398. There, the plaintiffs damages were caused by a projectile coming off a vehicle in highway traffic. The Supreme Court of Canada concluded that the general liability policy covers that portion of the loss attributable to non-

- 22 - auto-related negligence. We read into this that co-extensive causes will result in full coverage for the loss in most cases, even though it was not the intention of the underwriter to insure events taking place on Her Majesty s highways. In the case of automobile insurance, both for first-party coverage and for liability, the courts have taken a liberal approach in bringing losses within the ambit of coverage arising out of the ownership, use or operation of an automobile: Amos v. Insurance Corp. of British Columbia, [1995] 3 S.C.R. 405. In contrast, judicial consideration of the automobile exclusion in general liability policies has been rather strict, akin to the application of the limitation period under the Highway Traffic Act, when it was shorter than the general limitation period. In those instances, it was possible for an injured party to obtain the benefit of first-party or statutory accident benefits, while at the same time the accident did not arise from an automobile accident, for the purposes of a tort claim: see Re Ford-Smith Machine Co. (Bankruptcy), 2005 CanLII 9671 (Ont. S.C.J.), at paras. 25-30. Whatever your client s perspective on this issue, the Derksen principle should not be ignored in any instance where an accident has been caused by concurrent perils, even if one of them triggers a liability insurance exclusion. Why Coverage Matters All tort litigation lawyers are insurance lawyers, or must become insurance lawyers. Despite the flaw in that syllogism, the reality is that in tough cases, insurance can make the difference between a viable lawsuit and one which is not worth pursuing, for lack of a source of recovery. Insurance replaces an individual or corporate defendant decision-maker

- 23 - with an institutional risk manager. Insurance is a true game-changer in any litigation. While there is no one-size-fits-all answer to the consideration of liability insurance exclusions, there is no real substitute for careful analysis of the facts as they apply to insurance policy wordings, at every stage of a civil action. If you jump to conclusions, take the time to make sure they re the right ones. 2011 R. Lee Akazaki, C.S., Partner, Gilbertson Davis Emerson LLP; Past-President, Ontario Bar Association. Practices insurance litigation of all kinds, including professional responsibility and construction losses; the content of this paper is for professional education purposes only and not intended to provide a legal opinion. Contact: lakazaki@gilbertsondavis.com Blog: leeakazaki.com